ThinkVid: Benchmarking Visual Reasoning in Video Generative Models

Jun 1, 2025·

Wenjie Shu

· 1 min read

Abstract

A benchmark for evaluating visual reasoning ability of video generative models.

Type

Publication

Under review

Under review.

Last updated on Jun 1, 2025

Video Generation Evaluation Reasoning

Wenjie Shu

Authors

Wenjie Shu (he/him)

Research Asistant

← CoRe-GRPO: Consensus-driven and Region-focused RL for Human-Centric Generation in Lightweight Diffusion Models Jun 1, 2025

Temporal Regularization Makes Your Video Generator Stronger Mar 15, 2025 →